pandakraut

Sorry if I came across as criticizing the play style. Was just trying to point out that it's a different approach and that in your case medicine was probably the right answer. If I played real time I probably would have sworn off ever using skirmishers or cavalry ever again, they just love to get themselves killed. If you can keep the casualties down in the side battles I think you still have a good shot at making the campaign a success. You're just getting more politics at this point so the recruit situation should ease up a bit.

I've been able to float a lot of recruits and rifles on Legendary going the medicine first route as well. I think it may have more to do with playstyle than anything else. Heavy micromanagement of snipers, artillery, cavalry and falling back to avoid melee can result in very high kill ratios for far less casualties than in Lava's campaign. With that approach probably either politics or medicine can be successful. With the higher casualties that Lava's style generates my guess would be that medicine outpaces politics. In retrospect maxing or at least putting several points in politics before maxing econ might have been the better choice though. Someone could probably do the math to get actual numbers on politics/medicine/econ/training in terms of total money saved, but that's a ton of work for not much benefit.

Now if only I could follow my own advice. I always feel like I am running out of time and rush things especially when units are on the run.

If you can, try and take advantage of the timers as much as possible. Many battles, especially early on give you way more time than you need. Instead of constantly pushing encirclements you can hold positions in good cover, rest up for a few minutes and allow your artillery to catch up, and then continue the push at full strength. I usually keep my casualties down by avoiding melee as much as possible and making heavy use of artillery and snipers, but that also requires a lot of pausing so it's not really helpful with your play style. I usually completely ignore politics until medicine, economy, and training are maxed out. I'll get 3-5 in economy and training before starting medicine. I don't take enough casualties to need the recruits and economy/training/medicine make up the money difference. AO goes fairly high early on since I play with scaling manipulation, so it might not work as well with larger unit sizes.

@Hussar91 What difficulty were your capture results from? I've definitely gotten spencers from that mission, but I've also gotten worse weapons like you experienced. I've never seen anything close to an 85% rate of return of weapons when capturing units. You can definitely get whitworths and such from skirmishers, they are just somewhat rare and the numbers are small enough you only get very few back. To actually comment on Lava's playthrough instead of just debating mechanics, impressive work at Chancellorsville and at keeping the ai size down overall. I didn't realize it was possible to keep it that low with an army your size. It does look like you're starting to outpace your supply of recruits though. Look forward to seeing if you've snowballed things enough.

Didn't the weapon capture rate get changed at some point so that mass capturing wasn't overly profitable? Given the recruit capture cap, if you can safely reactivate surrendered units to farm more xp and kills off them it's probably the optimal choice either way. @quicksabre Once a unit is recaptured they are in the same state that they were captured in. But as long as you wipe them out afterwords the benefit should be the same or better(if someone can prove that there is a return exchange of troops).

I absolutely agree with your last statement about the scripting. I'm more skeptical of your theory about the impact of wiping out a smaller vs larger force is accurate, but I only have my own anecdotal data to work off of. Unfortunately getting actual data to confirm one way or the other involves a lot of time consuming testing. My current thought would be to use first bull run as a starting point. Fight the battle mostly as normal but with the goal of leaving as many enemy units unshattered as possible and clumped together. Make a save then finish the battle with all units left unshattered, half of enemy units shattered, and all units shattered. Then compare weapons and training as well as the actual unit sizes(using a recon 10 mod) to see if there is any noticeable difference. Enemy casualties should be fairly close on all 3 tests so if the unit destruction is important then hopefully this should show major differences. Performing a more extended test across multiple battles or a set of side battles and major battle would probably give a better set of data, but would be considerably more time intensive. The types of reinforcements are definitely semi-random if not completely random so repeat testing across a series of battles would be difficult. I'll see if I can get to the bull run test sometime soon.

If the most important thing is wiping out enemy brigades, then wouldn't reducing the enemy size to make wiping out enemy brigades easier be a good thing? Assuming you are interested in manipulating scaling of course.

For the Union you'd want to use Mule Shoe(1.75 million), Cold Harbor Day 1(1.5 million), or Marye's Heights(1.5 million). For the CSA Cold Harbor Day 1 allows 950k.

In my opinion the goal of building a large army while managing the enemy army size, training, and weapons is independent of utilizing techniques to manipulate scaling. I'd describe my approach to campaigns where I have used scaling techniques in almost the same way as you have above. Either way, my goal wasn't to turn this into another scaling discussion so I'm looking forward to seeing how your campaign progresses. To go in a different direction, if you're not already aware, reinforcement types and numbers are somewhat randomized. If the AISize drops below a minimum amount relative to the timeline of the campaign the reinforcement size will be increased to reach that minimum. So if you are not below the minimum it is possible to replay battles multiple times with similar kill rates and get different results. Might also be possible above the minimum but that is harder to test. Here are some test results from a dozen instant retreats from Shiloh on csa legendary. Small sample size but 1 and 2 were significantly more common. 3 only occurred once. Pre Battle Intelligence: 66-71 28-33 33-38 Random Result 1: 81-86 29-34 33-38 14700 fresh Random Result 2: 85-90 31-36 33-38 6400 well trained 12400 fresh Random Result 3: 79-84 29-34 46-51 weapons 12800 recruits Post Fighting Battle Intelligence: 50-55 33-38 33-38 6400 well trained 16400 transfer(unclear if recruits from message text)

Maybe I'm missing something, but if you are planning for a major battle by creating 1k sized units and then going into a side battle where you field units that are larger that's going to have a similar effect to what ballast units do. I've never tried always bringing the maximum number of brigades to every major battle though which is why I'm interested to see how scaling reacts to that(even if that's not your intention the data is still useful). Even on legendary, with high kill counts in the battles as the union it seems to be possible to get into situations where the CSA just doesn't have enough troops to adequately fight the battles.

Interesting approach on dealing with scaling. Looking forward to seeing how it works out long term. The voice audio was a little rough in some of the early episodes, but the last few videos have sounded good.

Just wanted to add that the next release will include perk tool tips and some other usability improvements so it should be much more convenient to try out.

If BG is getting that easy you may want to move up to MG. Scaling starts to become much more noticeable on MG and Legendary. I don't think it's really worth the effort of messing with below that. It's been demonstrated that you can completely ignore scaling if you want to and still be successful at the game. Now that it is somewhat known what affects scaling, deciding to utilize it to gain an advantage is just another tool available to the player if they choose to make use of it.

Some extra evidence for always being outnumbered 2:1 at Antietam as CSA. I recently tried a near minimum size army(1.1k - 1.3k man infantry brigades) into CSA Legendary Antietam. Scaling was pushed as low as it could go and I was still outnumbered 35k infantry to 70k. I would disagree somewhat with bringing more men always being an improvement. Bringing more men will result in enemy brigades being over sized. In my experience even if my brigades are smaller, it is easier to deal with 1.8k - 2k sized brigades than over sized ones. I also am able to field more brigades while using fewer recruits and buy fewer rifles. To actually describe how scaling works here is a summary from Aetius: "...For minor battles, you need to retain a lot of cheap weapons to build ballast brigades to reduce your average brigade size. For major battles, you should bring as many troops as you can up to the minimum enemy size (regardless of what they are armed with), and additional troops as you feel justified. Ideally, you want to fight minimum-size enemy forces with the largest forces you can muster. This limits your casualties and sharply reduces the number of free zombie troops the AI gets to replace losses, which eases some of the strategic problems with the AI having unlimited reinforcements." If you want more technical details: There are at least two scaling factors involved. The base army size will scale and separately each unit type will also scale. For example, if your army only has infantry units your total army size will cause non-infantry units to scale up as well. But the infantry units will scale much more significantly. Large amounts of cavalry, skirmishers, or artillery will have a lesser effect on total army size scaling but you'll encounter noticeably larger sizes of the corresponding unit type. There is also a minimum threshold for unit sizes that can drive scaling back up. These factors utilize the total number of men, the number of brigades, and the ratios of player units/men and ai players/men. Which gives you multiple factors to influence. Scaling also factors in allied units so battles involving those are harder to manipulate.

@LAVA Firearms and Efficiency are both used to generate multipliers in the damage calculation. Unlike the Weapon Accuracy values these are evaluated on an animation curve so their effect isn't really quantifiable unless you have access to the game code in Unity. Same for Damage Degradation.

The bonus per point of the career modifiers can be changed but at this point it's not possible to add a new bonus. In some situations the existing modifiers could be applied to a different in game calculation to give them an additional bonus. Politics is strong but depending on play style other options are higher priority for me.

The basic formula is weapon damage * (random value between accuracy low and accuracy high) * cover. Accuracy high is not currently displayed in game and cover will result in a damage reduction. There are other factors involved as well as a damage falloff as range to target increases. When evaluating weapons a rough estimate is multiplying the damage by the displayed accuracy and using that number to compare. Attacks will always hit so low accuracy high damage weapons can be useful.

I have not tested myself but I have heard that skirmisher efficiency starts dropping off after 250.

I haven't played many of the later battles where fortifications start to get useful using the perk changes. I pulled an old Laurel hill legendary save where my army was based on the weapon changes mod(sub-optimal rifles for vanilla and sub-optimal perk choices) and I played up to the counterattack portion of the battle and was able to hold with acceptable losses outnumbered ~25k to 54k. The allied units seemed to do fairly well in the fortifications. They would hold until 3-5 units piled up in front of them and routed them with rifle fire. Snipers to hit the flanks were very helpful in managing the AI buildup since the terrain is very favorable for it. If I tried again I would probably bring 3 units of them in my first deployment instead of bringing in more artillery. One thing that could be the issue is how the AI perk selections went. If you ended up against a bunch of max speed or max morale perk units that would be much more difficult to deal with. Both of those values are currently too strong and will be going down in the next version. If you want to post your save file I could give it a try with your setup. They are stored at C:\Users\<YourUserName>\AppData\LocalLow\Game Labs\Ultimate General Civil War\Save\CampaignBattle. The filenames aren't human readable so the easiest way to get the right one is just to load it up, make a new save, and sort by timestamp.

I haven't had any issues using detached skirmishers, though maybe I use them differently. I definitely don't expect them to stand up to fire for any period of time outside of cover with the unit sizes I'm using. I'd agree that somewhere between 1.6 and 1.8 is probably the range you want to shoot for if you want larger skirmishers. Good to know on the skirmishers efficiency. 250 is what I normally settle on in terms of available sniper weapons anyways. The extra weapons and recruits have never bothered me, but I also tend to horde resources just in case and then never use them. If you're using ballast units you always want at least 5-8k available to scale those up and down. I've never tried the max army size approach, but with anything less than that I think you will always eventually float recruits unless you are taking horrible losses.

While doing some unrelated testing I noticed that for Brandy Station and Supply Raid your army size is included in scaling even though only allied units deploy in the battle. Somewhat obvious in retrospect considering how side battles normally works, but I hadn't realized this before. Definitely don't forget to use ballast units prior to those battles as the allied units already make the scaling bad enough.

@aileycc Are your results from the weapons changes version or the perk changes version of the mod. Also, what difficulty are you playing on? Cav is definitely too good in both versions for different reasons and are being toned down in the next version. Snipers are amazing in both mod versions and in the normal game if you're willing to micromanage them. I would say they are close to mandatory on Legendary(but that could just be me). Mainly they allow you to deal large amounts of damage while taking few casualties. I have seen people use dismounted cav and non sniper skirmishers well but that's not something I've ever been successful with.

@LAVA There are at least two scaling factors involved. The base army size will scale and separately each unit type will also scale. For example, if your army only has infantry units your total army size will cause non-infantry units to scale up as well. But the infantry units will scale much more significantly. Large amounts of cavalry, skirmishers, or artillery will have a lesser effect on total army size scaling but you'll encounter noticeably larger sizes of the corresponding unit type. Two easy ways to see this is watching enemy guns decrease or increase as you add/remove infantry units and in some of the early battles using a cavalry heavy composition and seeing how large the cavalry units are in comparison to going infantry heavy. I have been playing with relatively small infantry units recently and using more cavalry. I have found that I need far fewer infantry ballast units but that I need far more artillery and cavalry ballast units. There also seems to be some kind of minimum threshold involving small unit sizes that can drive scaling back up. At Antietam CSA Leg I ended up fielding a third corps of small rookie units with muskets because fielding fewer men actually drove up scaling. I basically needed them anyways to hold the southern crossings and didn't have the rifles to add troops to my main units, but it was surprising to see removing units increase scaling in a major battle. There is probably some minimum size to unit number ratio that gets enforced but never comes up on larger unit sizes. @Grimthaur The sweet spot for minimum scaling seems to be somewhere between 1k and 1.3k depending on other factors. I've found that while ballast units are still beneficial they are far less impactfull then when I was using 1.5 to 2k sized brigades. So if you wanted to avoid using them and don't want to deal with over scaling that's probably the range to go for. These numbers are purely anecdotal but < 400 for cavalry, < 300 for skirmishers and <10 guns for artillery seems to be around the right spot as well. The side advantages of this in terms of fewer weapons and veterans needed is also very nice. I was able to invest the money I saved on rifles into veterans so post Antietam I had two full corps of 2+ star units. The rifles you can get through reputation are arguably a little too strong when playing on smaller unit sizes.

Unit stats always cap at 100. Having bonus stats from perks + unit stats that total to more than 100 just allows you to add rookies without a drop in those stats. I'm not sure if the units stats continue to increase to 100 even after the cap is hit with the perk bonus.